

This subdirectory contains the replication materials for the missing data sensitivity analyses we report in the 
paper.  Here we test for whether the causal effect estimates change when we impose first a prior that the 
missing data come from an extreme liberal distribution, and then second from an extreme conservative distribution.

To understand how to run the replications in this directory, you will first need to follow the directions for
replication in the mainmodel subdirectory.  The replications in this subdirectory are identical to the 
mainmodel, and the only difference is that the models in this subdirectory have the different priors for
the missing data distributions.  


